CLI
Work with your Nuvelics organization from the command line, in scripts, or from an AI agent.
The nuvelics command-line tool reaches your organization through the Nuvelics API. It suits two kinds of
work: scripting and automation by an integrator or a newsroom engineer, and unattended use by an AI agent
that can run shell commands. What you can do in the web app you can mostly do here too: list and approve
pitches, read stories and their outputs, manage data sources, newsdesks and credentials, and push finished
outputs on.
Every command works for one organization at a time, as the person or service key that signed in. It never does more than that credential is allowed to do.
Where to start
Read the pages in order the first time:
- Install says what is available today and what the tool needs on your machine.
- Sign in connects the tool to your organization and checks that it worked.
- Contexts explains the named profiles that decide which organization a command acts on.
- Output covers the output formats, and which to choose for people and for scripts.
- Filters and paging shows how to narrow a list and walk through all of it.
- Exit codes and errors lists every failure the tool can report and which ones are worth retrying.
- Scripts and agents puts it together for unattended use.
When you know what you want to run, the command reference has a page for every command, with its options.
If you want an AI assistant to work with Nuvelics through a chat client rather than a shell, see Agents. To call the API directly, see the API.